I'm not a developer, but I deal with different types of systems, and appreciate both Linux and Solaris for their respective strengths. In the telecom space, for instance, Solaris is well respected for building embedded applications. While AT&T invented Unix, they never meant it for critcal "five nines" real-time telephone call processing. Yet the dial tone on my desk comes from a Solaris-driven central office switch. (Not Lucent!) While the switch vendor's own code has crashed, the Solaris layer beneath takes a lickin' and keeps on tickin'.
I think the big fallacy in Linux is the driver ABI. Linus likes to change it, as a way of forcing hardware developers to have open-source drivers. Nice Stallmanesque politics, but impractical in the real world, for at least two different reasons.
1) Not all drivers can expose the source. This is often because complex devices hide proprietary details in the code. nVidia does that with its "compile in the stub" 3D drivers. Even more limiting are the wireless-card drivers, wherein regulatory approval is dependent on limiting user access to some of the chip registers which, in an open-source driver, could be used to create out-of-band or over-power emission. Life ain't all Ethernet cards nowadays. I had No Fun trying to make a PCI wireless card work with Linux, partially because of the (older) version dependency of the vendor's binary-only driver. Solaris and indeed most (not all) Microsoft OS versions have been better about that.
2) There's a lot of custom hardware out there. Sure, Linux users generally think about "computers" that are either "desktop" or "server" systems. But embedded systems are even more common. Solaris works in a lot of big ones, like aforementioned telephone switch. Some of those systems use different makers' boards; said phone switch, for instance, is made by a company that buys critical boards from other companies. Changes in the ABI would make a difficult revision process even harder. And even if you make your own peripherals, having to recompile or, gag, rewrite the drivers to meet Linux' latest idea of an ABI is, well, a serious pain in the kiester. Very unprofessional!
So while most mainstream dekstops do get better support in Linux, in part because of the better volume of applications, the Solaris approach still wins for those big systems where an hour of downtime is worth tens of thousands of dollars.
I'm sorry I'm not moderator today or I'd mod you up myself! Another worthless article about a worthless protocol, IPv6, breakfast of idiots.
Ridiculously-long addresses cause more packet overhead, since lots of packets carry nothing but ack's or short requests for data. And use VoIP, which needs small packets to control latency, and IPv6 makes the bandwidth preposterous. The usual excuse is that bandwidth is cheap, but that's not always true, especially if Uncle Sam or mommy isn't paying for it.
But the worthless article has even more wrong. It repeats the false claim, which by now has become a Big Lie of Bushian proportions, that IPv6 helps QoS. It doesn't do jack squat for QoS! You have one DiffServ bit in v4 already, and MPLS underneath IP for when things get more critical. (The two are sometimes used in tandem.) IPv6 has a useless flow indicator in the header. It's too short to have global significance, and IP is connectionless, so it does zilch. Just a waste of bits. Likewise, V6 doesn't do jack for security, since IPsec works with V4 already.
Nor does V6 help control router table sizes as the article claims. Sure, it's possible that existing networks, when renumbered in v6, can be more rationally aggregated. But a bigger address space invites more address blocks.
Look, I was around when V6 was invented. It was during the reign of the first President Bush. The Internet was not yet public. The IETF had pretty much settled on TUBA as the evolution of IP. TUBA was based on OSI CLNP, which was quite rational. (It wasn't all of CLNP, which had a zillion options; rather, it was a profile of CLNP.) The competition came from two half-baked undergraduate-quality proposals. Steve Deering, who was popular in IETF by then for doing some other good work, came up with "SIP" (Steve's IP). Paul Francis (Tsuchiya), another Nice Guy, came up with "PIP". They both played Yugo to TUBA's Mercedes. But "OSI" was unpopular, for religious reasons, and the taint wore off onto CLNP. At the last minute, IPv6 was created by merging some of PIP into SIP. And Vint Cerf, the Chauncey Gardner of the Internet, changed his IAB vote from TUBA to IPv6.
It's been about a bakers' dozen years since then and IPv6 is still two years away from widespread acceptance. With any luck it'll stay that far away. It's useless and costly. IPv6 is just an ego boost for some tech weenie poseurs, and a scheme by Cisco and its ilk to extract more money for new hardware that will be needed to cope with this unnecessary hack.
The ARPAnet, not quite the Internet
on
The Internet At 35
·
· Score: 5, Informative
Well, let's not let the title get too carried away on accuracy, even disregarding its subtraction error. In 1969, the prototype ARPAnet started up. It used NCP (TCP/IP came later). It didn't become the "Internet" until there were multiple interconnected networks, and that was not until the early 1980s, after the TCP/IP transition (which was completed in 1983). There were multiple networks once the more production-oriented MILNET split off of the more research-y ARPAnet. And after that came CSnet and all sorts of others.
But yes, it was in many ways better in the early days (pre-1993), because there was no spam, or for that matter any other advertising. Although Google and the like do sort of make up for it.
SCO's claim against AIX based on the inclusions of SVR5 code is only valid if it's SCO's to license. While SCO does have the right to issue SVR4 licenses, it seems quite probable, and is likely to be made final by a court soon, that SCO does not/i> own SVR4 copyrights at all. Novell retained them. So if there is an aggrieved party with a cause of action against IBM, it's Novell. But if Novell decides that there's no problem, then SCO is SOL.
And that of course assumes that SCO's claims are otherwise valid, which is fairly unlikely given their record of mendacity.
As the previous poster noted, Iridium is still in business. The original owner, Iridium LLC, went kaput after spending about $5B, but the constellation was picked up for about a penny on the dollar by "Iridium Satelitte", a different company, and they're keeping it going quite nicely. Uncle Sam uses it a lot.
But Iridium's bandwidth is very low -- about 2400 bps. Low-earth-orbit satellites have less latency, of course, but the cost of bandwidth turns out to be a problem. Especially if you have to pay full price for them, vs. getting them as bankruptcy assets. Geostationary satellite turns out to be cheaper.
I couldn't tell you what the learning curve of GRASS is like -- it's so steep up front that I've never been able to get anywhere. I've installed it, but even using the Tcl menus, I haven't been able to really use it. I think it takes having an expert sitting next to me showing me what to do. I don't have such an expert here.
To be sure, I did have such help to get started in MapInfo, but once I got the hang of it (quickly), the learning curve for most things was not too bad. Although it is rather sloppy in some respects.
The "briefing" has a good collection of pointers to open-source applications out there. But as a fan of the commercial Windows GIS product MapInfo, I am frustrated by the lack of an open source alternative, and by the lack of comparable tools for Linux. GRASS is pretty powerful, but it's not something anybody can just start using; it's more like something a Unix GIS professional (difficult but powerful systems like ESRIs) would find interesting.
This note from the briefing is most telling:
Note: The saturated commercial market for cartography tools, the high level of effort to achieve a usable tools, and the appeal of other cutting edge projects have combined to deter any active development on user-friendly paper map production tools. As with the OpenOffice experience in Linux, it would probably require a dedicated multi-year funded project to produce a core product with sufficient technical mass that an open source community could reasonably continue with enhancements and support.
In other words, don't expect to find a complete open source end-user application within your lifetime.
This is, alas, common in the open source world. Everybody does their own toolkit that does 90% of what other toolkits do, adds 10% of its own, and assumes that the user is a person who gets their jollies from writing code, not actually using the application with production data.
Well, if you want to get picky, the name Sprint did not come from an acronym as you said it did; rather, the name "Sprint" was chosen from an employee "name the company" contest in the late 1970s. It was the Southern Pacific Communications Company at the time, but the SP railroad sold it shortly afterwards.
Someone earlier noted that in the UK, Energis had trouble with their easements when they used power easements to run telecom. In the US this is not usually a problem for utilities, but it has been a problem on some railway easements. Railroads have been leasing space to telecom companies for years, but some of their own rights-of-way may not have allowed it. This has been rather messy in a few places, as some landowners had to be placated (paid off) when their lawyers figured out that they could turn the screws.
BPL is of course an awful idea, for technical reasons that others have pointed out. I suspect however that intentional interference is part of the design. Short-wave radio, while unpopular in the United States, is still a valued form of getting past the censors in places like China and Arabia. BPL is probably being geared up for export to those places, where it can finally do the job that hundreds of cold-war-era jammers could not do. A broadband shortwave jammer in every neighborhood? Used to feed filtered, government-sanctioned content? A tyrant's dream!
Bush and his cronies in the increasingly-concentrated media empires couldn't have similar ideas, of course....
You misunderstood what happened. AT&T did not want sharing and resale. The FCC allowed them to maintain that restriction in their tariffs for some years. Then, in the mid-1970s, as the dawn of competition was breaking, the FCC ordered AT&T and the other telcos to drop that restriction. Left to their own, AT&T would have simply shut off service to anyone who did something they didn't like, like sharing. The FCC reined them in.
Most state regulators, and indeed, the FCC, are often mouthpieces for the industries they regulated. But there have been times when regulators did their job and actually regulated; that's when progress happens.
McCullagh's position on CNET is wrongheaded, and highly anticompetitive. His article actually cites Huber's book, which proposed converting existing radio station licenses into property, so that the licensee of an FM radio station instead ends up with chattel ownership of 200 kHz, to do what they want with it. It's a wingnut's fantasy, a huge transfer of public wealth (the radio spectrum) to private interests (licensees), with the current need to serve the "public interest" replaced by a total obeisance to shareholders' interests, in the name of doctrinaire laissez-faire capitalism. The current licensing system is obsolete, and the FCC's anti-indecensy crusade is nutty, but "property rights" just transfer the problem to courts that lack the FCC's technical staff expertise (some of which does still exist).
But it's the telecom area that really needs attention. Yes, the Powell FCC is profoundly broken. It regulates by indirection, picking winners and losers privately and coming up with indirect ways to favor them. Its main beneficiaries are the lawyers who try to pick up after them. So one might think that the FCC's charter is broken, but that's not it at all. It's simply the leadership and the politics behind it; this FCC, much worse than its predecessor, is clearly led by a celebrity princeling who just doesn't get it. A change in leadership is necessary, not abolition.
The reason is simply that the telecommunications industry is highly concentrated. The Incumbent Local Exchange Carriers have monopoly power. In the European Union, IIRC, a company with a 25% market share is suspected of having monopoly power, and scrutinized for abuses thereof. The USA is very, very loose on antitrust regulation, and the ILEC monopolies were granted legally, so the antitrust laws only (per the Supreme Court's recent Trinko decision) apply to attempts to extend the monopolies into new areas. Demonopolization is entirely the province of the Telecom Act, not antitrust. And the Telecom Act puts the FCC in the lead. Without regulation, a monopoly will simply squash competitors. This is particularly true in telecom for two reasons. One is the "natural monopoly". This refers to the case where a given industry has large economies of scale and a dominant provider. The unit cost of the dominant provider is thus lower than that of a new competitor, so the economics of competition are dismal.
The other reason is the network effect: A network's value rises with the number of users that it reaches. Federal regulations, enforced by the FCC, require *interconnection* between networks. A CLEC with ten customers can interconnect as a peer with the incumbent. The incumbent, of course, has no interest in allowing this. The incumbent, absent regulation, would shut off interconnection to its competitors in a heartbeat. This wouldn't occur if the incumbent's market share were small, but it's necessary to force interconnection *until* the monopoly is broken, and the ex-monopoly has a pecuniary interest in retaining interconnection.
The Internet has no dominant player, so everyone willingly interconnects. Worldcom wasn't allowed to buy Sprint, largely for that reason. In an FCC-less fully-deregulated world, Verizon and SBC would not be so kind. They might deign to permit competitors to purchase access to their networks, as premium-priced customers rather than peers, if they thought it was profitable enough. That's hardly a way to get competition though.
Remember, the only reason the public Internet exists is because the FCC, over the *strenuous* objections of the Bell System, overrode restrictions on "sharing" of leased lines. Before that, non-common-carrier networks (like the Internet) could not be run between customers. Leased lines, necessary for high-speed data, were limited to intra-company use. And the FCC, over the *strenuous* objections of ILECs nationwide, overrode restrictions on "foreign attachments", devices like modems, answering machines, telephone sets, and PBXs. Before 1
Good point about NT. The original author of Windows NT, and I think still a leader in MS OS development, was David N. Cutler. He joined MS from DEC. He left DEC when they cancelled Mica, a VMS-like operating system that was decoupled from VAX hardware. Mica was being written at DECwest near Seattle, a site opened at Cutler's request -- he was that influential.
Before moving west, Cutler had been the lead author of VAX/VMS (1977), reportely writing a rather large share of the V1.0 kernel himself. (He was said to be very, very fast at assembly programming, but created inefficient code. Very "V1".) Before writing VMS, he had worked on RSX-11. The original VAX/VMS was upward compatible from RSX, and indeed included RSX as a mode. (By the mid-1980s, PDP-11 emulation capabilities had been dropped from VAX hardware.)
People who are familiar with the innards have said that NT is closely modeled on VMS (which was decoupled from the VAX when Alpha came out; it has since been ported to Itanic). But that's not a reason to whine. Cutler had some good ideas, and he expressed them several different ways. He almost certainly didn't lift DEC code directly into NT -- although at the time, DEC did indeed make most VMS sources readily available (on microfiche -- not to be too easy to use). Oh yeah, VMS was mostly written in assembler and BLISS, not C.
If Linus' efforts were unoriginal, how do Cutler's at Microsoft compare? Is the pot calling the kettle black? Personally I think neither is guilty of wrongdoing. I'm even getting to sort of like Win2k, which I use on the laptop, though XP's bloat is just annoying, and 98's instability is really annoying.
No, no shame. SpecOps was trolling for dollars by keeping quiet about the WINE code they're using. Now they're spinning.
So in this new letter, they're admitting that there's WINE in there, though not saying how much. And they're adding improvements atop it, which Codeweavers also does. And when it's released, per the GPL, they'll apparently release their modified source code where required. Okay. Whether David is useful or not remains to be seen.
You're not reading the OP correctly. My business, like many small businesses, doesn't own a mail server. It owns a domain name, which works via a mail-forwarding service on mydomain. Outgoing mail goes via whatever ISP I happen to be on at the time; i.e., the one at home when I'm home, the one at the office when I'm at the office, or the one at the hotel when I'm at a hotel.
My clients don't care where my mail originated. They know it's from me. I'm not Paypal or BankOne.
If I had a mail server of my own, then another popular (but dumb) antispam measure, port 25 blocking, would cause it to be inaccessible from at least some locations.
Now it sounds like a bad idea for both semantic (what it does) and syntactic (how it is coded) reasons!
The syntactic bit is easy -- XML is hardly appropriate for a DNS function. Mickeysoft is running around patenting XML schemas, and it adds a new layer of complexity to DNS. But then bad syntax is usually dealt with by code.
The semantic bit is worse -- SPF doesn't block spam unless the mail system makes it mandatory, after all, so until 100% compliance is reached, non-SPF mail will still have to be accepted. But wait -- SPF doesn't block spam! It just blocks spam where the From: is not right. Spammers can still create new domains on a hit-and-run basis, and they'll pass SPF. So it's another blast-proof vault door stuck onto a grass hut, a silly waste of time. The only potential real benefit, I suspect, would be to make phishing harder. The address will have to be slightly different from the spoofed domain. But that leaves plenty of opportunity to create deceptively-close hit-and-run domains (like, say, pay-pa1-approva1.com).
Worse, of course, is the collateral damage. How will I be able to send mail using my own business' domain, as I do today, when it is going out via an ISP server? My "from" address is an alias, not a real sender, and I use it to send via more than one ISP, depending on where I am. SPF seems to make this a lot harder, thereby forcing more people to put their ISPs' name in the From: field, rather than their own. Since email is not portable, a user's address is lost when they change ISPs, or when their ISP changes names (mediaone->attbi->comcast). Personal domains (forwarded via a service like mydomain) solve this. Will SPF kill mydomain?
I repeat what I've said before. The only way to kill spam is to stop having all email be totally, absolutely, "free" of charge in any quantity. This is not the topic to discuss solutions, but they are certainly possible, and they aren't SPF.
Allowing "new competition to enter the arena" is often ridiculous, when there is a strong natural monopoly component to the service in question. Natural monopoly is an economic term for a product with a very high entry cost to produce and a large economy of scale. Telephone wire is a good example: If somebody has a 100% market share, and loses 20% to a competitor, and the cost is to run the wire past the customer site (which is the case), then the incumbent's unit cost will be 1/4 of the competitor's. This makes competition extremely difficult.
The Telecom Act recognized this, by requiring the incumbents to unbundle their facilities. The incumbents went along with the deal (getting other things in trade), but immediately tried to back out. It's bloody out there now on the telecom streets.
SBC, VZ, and the other incumbents have a natural monopoly on the local wire, even though it's no longer a de jure monopoly. There's no natural monopoly on Internet service, servers, information, etc., but if the monopolists can exert full control on content, as Gilder wants and MCI doesn't want them to, then all of those competitors will be shut out.
The original post is wrong. It implies that Gilder supports the MCI position. But Gilder's testimony opposes it. Gilder is, like many extreme right-wingers, a fan of unregulated monopolies. He does not believe that monopolies should be regulated; he thinks that technology will magically render them powerless. His record in picking technology investments, however, is rather spotty, to be charitable.
In Gilder's world, the incumbent telephone companies, who had government-granted monopolies, should be allowed to have total control over their usage and content. No competing ISPs, no uncensored web sites, no competing web merchants, if that's what the monopoly wants. If they want to charge $100/month plus $5/hour for dial-up access, fine. Anything else is, to him, excessive government regulation. He'd permit somebody else, of course, to string new wires on the street, but the impracticality of that is not his problem.
He should be relegated to the dumpster of other failed nut case prophets, and left with his fellow creationists to ponder the problems of a world where scientists and rational thinkers are allowed to question his faith.
Intel licensed AMD to produce their designs, as a second source, up through the 80286. Intel masks and all. By the time the 386 came out, Intel didn't need AMD any more (they had multiple fabs and a good enough reputation, plus a lock on PC-compatible chips). So they told AMD that the agreement didn't apply any more. I don't remember if AMD won or lost on the 80386. But it certainly didn't last until the 486. So AMD did their own design, without any help from Intel. The court did note that a number could not be trademarked. It was thus never the "80486"; I think "i486" was a trademark, not that anybody cared, and that's why the next Intel chip was "Pentium".
AMD's "586"-class chip, the K5, was a dog. They then bought NexGen and adapted its RISC-innard design to the K6, which rocked, and fit a Pentium socket. Intel put tighter patents on the PII socket so AMD built the Athlon on DEC's Alpha socket electrical design.
Intel didn't have to change the ISA (drop the NX, for instance) in order to be legal. Either they goofed, or they sabotaged their own 64-bit x86 upgrade (as others here have suggested) in order to create a niche for the Itanic.
Lexmark obviously makes the Xerox printer in question. They design in little differences, like that ink clip, in order to create separate ink streams, which are where the money is. The article shows how one person got around one such difference.
Dell printers are also Lexmark; I don't know if they differentiate on ink. But if you read the printer reviews in the PC mags, you'll see the estimated per-page ink prices. Guess whose are highest? Lexmark ink is much costlier than the other major brands (HP, Canon, Epson), with Dell-branded ink usually a notch higher yet.
Good point -- Mandrake ADDS VALUE via its own proprietary printer-installation routine. Maybe it's 99% CUPS, maybe it's 99% their own, but at least I got it to work. Fedora Core sounds like it's running raw CUPS. I couldn't imagine Red Hat going out of their way to make it that bad!
And Kudos to ESR on this piece in general. CUPS is just one example of the abuse of GUIs across Linux. The X Windows configuration tools in some distros are even worse. When you use a GUI that sets some value or other, the first rule is that the value displayed should be the CURRENT value. After all, it's a tool for seeing, as well as setting. But no, in some Linux tools, the GUI always comes up with some weird default value! It ignores what's there and wants ohhh so badly to overwrite it with something else. What an awful thing to do!
That's PRECISELY the kind of UI polishing that Linux needs. Not a few more sci-fi desktop themes.
AMD's big problem isn't marketing, it's manufacturing. AMD is doing a pretty good job with the resources they have, but Intel has much larger capacity. Nobody fabs like Intel, and nobody has fabs like Intel. So if AMD had more orders, they'd just have bigger backlogs. They may be able to get some third-party fabrication, like from IBM, but Intel would not let its fabs sit idle. I prefer AMD chips myself, at least for desktop-type applications (Intel still wins on mobile chips), but this is not like OS/2 (which I also preferred in its heyday) -- chipmaking is expensive!
The article, which was truly stupid, ran on ZDNet. While that sounds like Ziff-Davis, it's not. ZDNet is part of CNET, a separate company; they bough it from Ziff-Davis several years ago. Ziff-Davis still puts out PC Magazine (pcmag.com) and Extreme Tech, among other things. Sort of like how Lycos owns Wired.com, not Wired magazine, though those two still have a working relationship that is not apparent between ZDNet and CNET.
Among CNET properties, ZDNet seems to be oriented towards Windows-loving IT pros. Maybe that's part of its Ziff-Davis heritage, or just a marketing decision.
I've posted this before, when the topic came up on another occasion, but it's worth repeating.
The FCC is NOT going to regulate computer-to-computer "phone" calls. If you run voice over your Internet connection, as an application, it's your business, and that's that. Even the guy who drafted the infamous ACTA petition in 1996 now thinks VoIP is cool stuff.
The problem is the phone call between the consumer with a plain old phone line and the VoIP network. "Phone to phone" and "phone to computer" calls have a telco leg that's just a plain old voice call. Under current law, a phone call can be either "telephone exchange service" or "exchange access service". The former is basically taken to mean a local call, though the legal definition is a bit more expansive. The latter is taken to be the local phone company's leg of a toll call (what AT&T or MCI buys). Guess which one costs more.
Now if all VoIP calls were treated as local ("telephone exchange service"), then the local telephone companies (think: Bells) would lose money that they now make from exchange access service ("switched access"). And the rural phone companies, who charge the long distance companies MUCH more than the Bells for that service, in order to compensate for higher costs (that is, to subsidize local service to the sticks), are very protective of switched access revenues. And the flyover states each have two senators.
So the main issue will come up around the far end of a Vonage call, for instance -- if Vonage is a long-distance company, they will have to pay access when they deliver a long distance call. Just like other long distance companies. Skype's on-net calls, and FWD, won't be touched as long as they are on net. Count on it.
Ideally, the whole access thing would go away, and the distinction between access and local would be moot. That's the way it works in msot of Europe, I think -- it's an American tradition to classify things to death, and let the lawyers litigate like crazy over the classification. How many billable lawyer hours do you think this case will be worth in Washington?
Those are interesting, thank you for posting them.
Both, however, seem dependent upon having both sides of the conversation join. In the case of Mailbox Reputation Network, it seems to be a voucher-based (as in I vouch for you, a web of trust) system. In the case of sudonames, it seems to be a whitelist system, where non-members can't send mail to you.
If both were universal, they'd probably work, especially sudonames, because that seems to count the credits, making host-hijacking worthless (it would run out of credits, which wouldn't happen right away with MRN, as I see it). But if I joined sudonames today, random people trying to reach my sudonames address would get bounced.
But then any micropostage scheme could have that problem, even mine, though a whitelist would exempt recipients from needing credits (stampettes).
I'm drowning in spam, and it's getting in the way of my job. The only solution that can possibly work is one that involves putting a price tag on spam. So here's my proposal (which I've put on here before, btw; this is not a new topic). The only way to put a price tag on spam is to put a price tag on email. But it doesn't have to apply to all email.
The price, then, is for the right to touch MY mailbox IF you're a stranger -- if you're a mailing list that I've subscribed to, you would go onto my whitelist, and come in postage-free. If you are somebody I know, you go onto my whitelist, and come in postage-free. Yes, for this to work, there has to be some way for the POP server (NOT the client) to maintain per-user whitelists.
If you're not on my whitelist, you need to use a one-time "stampette", whose price would have to be high enough to discourage spammers, but low enough to not bother anybody worthwhile. I'm thinking around a quarter-cent per message, but it wouldn't be fixed by anyone in particular. These stampettes would be issued on a free-market basis, and anyone could set up a micropostage service, provided that the *recipient* whitelisted it. So if somebody were giving away stamps at, oh, a million per dollar, then spammers would use them, and those stamps wouldn't be on my whitelist. Again, it's a free market solution, no government intervention.
ISPs, in this scheme, should issue all subscribers a batch of stampettes (which mail clients would learn quickly to attach, if needed). A thousand for a quarter-dollar (or quarter-Euro) would be more than enough for a month, don't you think? How many strangers (or first-time correspondents) do you write to?
The point is: IPv4 is not the problem; IPv6 is not a good solution to the alleged problems.
Not all of the IPv4 address space has been parceled out. ICANN has a lot of the original "Class A" space available. The space from 65-126 was never allocated as Class A; it has been parceled out to the three worldwide number assignment bodies on a demand basis. Quite a bit is left. There are also some Class As that can probably still be reclaimed, in whole or in part. NAT has also helped a lot in holding down demand for numbers.
Even if numbers were in desperately short supply and IPv4 couldn't handle the job, IPv6 wouldn't be the answer. It's plug-ugly, the bastard child of two amateurish hacks by IETF insiders (Steve Deering's SIP -- the current SIP is at least the second holder of the name -- and Paul Francis/Tsuchiya's PIP), melded together sloppily in order to get "consensus". IAB had already accepted TUBA, a far cleaner solution, but Vint changed his vote. What a friggin' disaster. TUBA (TCP and UDP over a CLNP profile) had already been implemented on all of the major routers of the day. It just hit a wall of "NIH", since its creation was tained by its OSI connection.
But that was all before the Internet was big or open. If a replacement for v4 were really needed, it should not be yet another old hack. It should be something built with today's requirements in mind, not 1990's. Real network research, alas, seems to have shut down at about the time that the Internet became commercially important. Too valuable to question, I suppose, and all the newcomers from Microsoft to SCO must imagine that it must have been well thought out in the first place (hah! it was government research, still alpha or maybe early beta work in progress) but that is a terrible way to maintain it.
And back to our Asian friends: Software has never been their strong point. Nor has questioning authority; too many, I suspect, assume that the TCP/IP suite has too much authority behind it. Asia's marvelous at mass-producing hardware, an art which involves being able to reproduce things perfecty in media that make it difficult to do so. So if they accept IPv6, it is not necessarily proof that it's, say, Toyota-grade technology. Even Japan has its clunkers. Remember Pink Lady and Jeff?)
Slashdot Mirror
I'm not a developer, but I deal with different types of systems, and appreciate both Linux and Solaris for their respective strengths. In the telecom space, for instance, Solaris is well respected for building embedded applications. While AT&T invented Unix, they never meant it for critcal "five nines" real-time telephone call processing. Yet the dial tone on my desk comes from a Solaris-driven central office switch. (Not Lucent!) While the switch vendor's own code has crashed, the Solaris layer beneath takes a lickin' and keeps on tickin'.
I think the big fallacy in Linux is the driver ABI. Linus likes to change it, as a way of forcing hardware developers to have open-source drivers. Nice Stallmanesque politics, but impractical in the real world, for at least two different reasons.
1) Not all drivers can expose the source. This is often because complex devices hide proprietary details in the code. nVidia does that with its "compile in the stub" 3D drivers. Even more limiting are the wireless-card drivers, wherein regulatory approval is dependent on limiting user access to some of the chip registers which, in an open-source driver, could be used to create out-of-band or over-power emission. Life ain't all Ethernet cards nowadays. I had No Fun trying to make a PCI wireless card work with Linux, partially because of the (older) version dependency of the vendor's binary-only driver. Solaris and indeed most (not all) Microsoft OS versions have been better about that.
2) There's a lot of custom hardware out there. Sure, Linux users generally think about "computers" that are either "desktop" or "server" systems. But embedded systems are even more common. Solaris works in a lot of big ones, like aforementioned telephone switch. Some of those systems use different makers' boards; said phone switch, for instance, is made by a company that buys critical boards from other companies. Changes in the ABI would make a difficult revision process even harder. And even if you make your own peripherals, having to recompile or, gag, rewrite the drivers to meet Linux' latest idea of an ABI is, well, a serious pain in the kiester. Very unprofessional!
So while most mainstream dekstops do get better support in Linux, in part because of the better volume of applications, the Solaris approach still wins for those big systems where an hour of downtime is worth tens of thousands of dollars.
I'm sorry I'm not moderator today or I'd mod you up myself! Another worthless article about a worthless protocol, IPv6, breakfast of idiots.
Ridiculously-long addresses cause more packet overhead, since lots of packets carry nothing but ack's or short requests for data. And use VoIP, which needs small packets to control latency, and IPv6 makes the bandwidth preposterous. The usual excuse is that bandwidth is cheap, but that's not always true, especially if Uncle Sam or mommy isn't paying for it.
But the worthless article has even more wrong. It repeats the false claim, which by now has become a Big Lie of Bushian proportions, that IPv6 helps QoS. It doesn't do jack squat for QoS! You have one DiffServ bit in v4 already, and MPLS underneath IP for when things get more critical. (The two are sometimes used in tandem.) IPv6 has a useless flow indicator in the header. It's too short to have global significance, and IP is connectionless, so it does zilch. Just a waste of bits. Likewise, V6 doesn't do jack for security, since IPsec works with V4 already.
Nor does V6 help control router table sizes as the article claims. Sure, it's possible that existing networks, when renumbered in v6, can be more rationally aggregated. But a bigger address space invites more address blocks.
Look, I was around when V6 was invented. It was during the reign of the first President Bush. The Internet was not yet public. The IETF had pretty much settled on TUBA as the evolution of IP. TUBA was based on OSI CLNP, which was quite rational. (It wasn't all of CLNP, which had a zillion options; rather, it was a profile of CLNP.) The competition came from two half-baked undergraduate-quality proposals. Steve Deering, who was popular in IETF by then for doing some other good work, came up with "SIP" (Steve's IP). Paul Francis (Tsuchiya), another Nice Guy, came up with "PIP". They both played Yugo to TUBA's Mercedes. But "OSI" was unpopular, for religious reasons, and the taint wore off onto CLNP. At the last minute, IPv6 was created by merging some of PIP into SIP. And Vint Cerf, the Chauncey Gardner of the Internet, changed his IAB vote from TUBA to IPv6.
It's been about a bakers' dozen years since then and IPv6 is still two years away from widespread acceptance. With any luck it'll stay that far away. It's useless and costly. IPv6 is just an ego boost for some tech weenie poseurs, and a scheme by Cisco and its ilk to extract more money for new hardware that will be needed to cope with this unnecessary hack.
Well, let's not let the title get too carried away on accuracy, even disregarding its subtraction error. In 1969, the prototype ARPAnet started up. It used NCP (TCP/IP came later). It didn't become the "Internet" until there were multiple interconnected networks, and that was not until the early 1980s, after the TCP/IP transition (which was completed in 1983). There were multiple networks once the more production-oriented MILNET split off of the more research-y ARPAnet. And after that came CSnet and all sorts of others.
But yes, it was in many ways better in the early days (pre-1993), because there was no spam, or for that matter any other advertising. Although Google and the like do sort of make up for it.
SCO's claim against AIX based on the inclusions of SVR5 code is only valid if it's SCO's to license. While SCO does have the right to issue SVR4 licenses, it seems quite probable, and is likely to be made final by a court soon, that SCO does not/i> own SVR4 copyrights at all. Novell retained them. So if there is an aggrieved party with a cause of action against IBM, it's Novell. But if Novell decides that there's no problem, then SCO is SOL.
And that of course assumes that SCO's claims are otherwise valid, which is fairly unlikely given their record of mendacity.
As the previous poster noted, Iridium is still in business. The original owner, Iridium LLC, went kaput after spending about $5B, but the constellation was picked up for about a penny on the dollar by "Iridium Satelitte", a different company, and they're keeping it going quite nicely. Uncle Sam uses it a lot.
But Iridium's bandwidth is very low -- about 2400 bps. Low-earth-orbit satellites have less latency, of course, but the cost of bandwidth turns out to be a problem. Especially if you have to pay full price for them, vs. getting them as bankruptcy assets. Geostationary satellite turns out to be cheaper.
I couldn't tell you what the learning curve of GRASS is like -- it's so steep up front that I've never been able to get anywhere. I've installed it, but even using the Tcl menus, I haven't been able to really use it. I think it takes having an expert sitting next to me showing me what to do. I don't have such an expert here.
To be sure, I did have such help to get started in MapInfo, but once I got the hang of it (quickly), the learning curve for most things was not too bad. Although it is rather sloppy in some respects.
The "briefing" has a good collection of pointers to open-source applications out there. But as a fan of the commercial Windows GIS product MapInfo, I am frustrated by the lack of an open source alternative, and by the lack of comparable tools for Linux. GRASS is pretty powerful, but it's not something anybody can just start using; it's more like something a Unix GIS professional (difficult but powerful systems like ESRIs) would find interesting.
This note from the briefing is most telling:
Note: The saturated commercial market for cartography tools, the high level of effort to achieve a usable tools, and the appeal of other cutting edge projects have combined to deter any active development on user-friendly paper map production tools. As with the OpenOffice experience in Linux, it would probably require a dedicated multi-year funded project to produce a core product with sufficient technical mass that an open source community could reasonably continue with enhancements and support.
In other words, don't expect to find a complete open source end-user application within your lifetime.
This is, alas, common in the open source world. Everybody does their own toolkit that does 90% of what other toolkits do, adds 10% of its own, and assumes that the user is a person who gets their jollies from writing code, not actually using the application with production data.
Well, if you want to get picky, the name Sprint did not come from an acronym as you said it did; rather, the name "Sprint" was chosen from an employee "name the company" contest in the late 1970s. It was the Southern Pacific Communications Company at the time, but the SP railroad sold it shortly afterwards.
Someone earlier noted that in the UK, Energis had trouble with their easements when they used power easements to run telecom. In the US this is not usually a problem for utilities, but it has been a problem on some railway easements. Railroads have been leasing space to telecom companies for years, but some of their own rights-of-way may not have allowed it. This has been rather messy in a few places, as some landowners had to be placated (paid off) when their lawyers figured out that they could turn the screws.
BPL is of course an awful idea, for technical reasons that others have pointed out. I suspect however that intentional interference is part of the design. Short-wave radio, while unpopular in the United States, is still a valued form of getting past the censors in places like China and Arabia. BPL is probably being geared up for export to those places, where it can finally do the job that hundreds of cold-war-era jammers could not do. A broadband shortwave jammer in every neighborhood? Used to feed filtered, government-sanctioned content? A tyrant's dream!
Bush and his cronies in the increasingly-concentrated media empires couldn't have similar ideas, of course....
You misunderstood what happened. AT&T did not want sharing and resale. The FCC allowed them to maintain that restriction in their tariffs for some years. Then, in the mid-1970s, as the dawn of competition was breaking, the FCC ordered AT&T and the other telcos to drop that restriction. Left to their own, AT&T would have simply shut off service to anyone who did something they didn't like, like sharing. The FCC reined them in.
Most state regulators, and indeed, the FCC, are often mouthpieces for the industries they regulated. But there have been times when regulators did their job and actually regulated; that's when progress happens.
McCullagh's position on CNET is wrongheaded, and highly anticompetitive. His article actually cites Huber's book, which proposed converting existing radio station licenses into property, so that the licensee of an FM radio station instead ends up with chattel ownership of 200 kHz, to do what they want with it. It's a wingnut's fantasy, a huge transfer of public wealth (the radio spectrum) to private interests (licensees), with the current need to serve the "public interest" replaced by a total obeisance to shareholders' interests, in the name of doctrinaire laissez-faire capitalism. The current licensing system is obsolete, and the FCC's anti-indecensy crusade is nutty, but "property rights" just transfer the problem to courts that lack the FCC's technical staff expertise (some of which does still exist).
But it's the telecom area that really needs attention. Yes, the Powell FCC is profoundly broken. It regulates by indirection, picking winners and losers privately and coming up with indirect ways to favor them. Its main beneficiaries are the lawyers who try to pick up after them. So one might think that the FCC's charter is broken, but that's not it at all. It's simply the leadership and the politics behind it; this FCC, much worse than its predecessor, is clearly led by a celebrity princeling who just doesn't get it. A change in leadership is necessary, not abolition.
The reason is simply that the telecommunications industry is highly concentrated. The Incumbent Local Exchange Carriers have monopoly power. In the European Union, IIRC, a company with a 25% market share is suspected of having monopoly power, and scrutinized for abuses thereof. The USA is very, very loose on antitrust regulation, and the ILEC monopolies were granted legally, so the antitrust laws only (per the Supreme Court's recent Trinko decision) apply to attempts to extend the monopolies into new areas. Demonopolization is entirely the province of the Telecom Act, not antitrust. And the Telecom Act puts the FCC in the lead. Without regulation, a monopoly will simply squash competitors. This is particularly true in telecom for two reasons. One is the "natural monopoly". This refers to the case where a given industry has large economies of scale and a dominant provider. The unit cost of the dominant provider is thus lower than that of a new competitor, so the economics of competition are dismal.
The other reason is the network effect: A network's value rises with the number of users that it reaches. Federal regulations, enforced by the FCC, require *interconnection* between networks. A CLEC with ten customers can interconnect as a peer with the incumbent. The incumbent, of course, has no interest in allowing this. The incumbent, absent regulation, would shut off interconnection to its competitors in a heartbeat. This wouldn't occur if the incumbent's market share were small, but it's necessary to force interconnection *until* the monopoly is broken, and the ex-monopoly has a pecuniary interest in retaining interconnection.
The Internet has no dominant player, so everyone willingly interconnects. Worldcom wasn't allowed to buy Sprint, largely for that reason. In an FCC-less fully-deregulated world, Verizon and SBC would not be so kind. They might deign to permit competitors to purchase access to their networks, as premium-priced customers rather than peers, if they thought it was profitable enough. That's hardly a way to get competition though.
Remember, the only reason the public Internet exists is because the FCC, over the *strenuous* objections of the Bell System, overrode restrictions on "sharing" of leased lines. Before that, non-common-carrier networks (like the Internet) could not be run between customers. Leased lines, necessary for high-speed data, were limited to intra-company use. And the FCC, over the *strenuous* objections of ILECs nationwide, overrode restrictions on "foreign attachments", devices like modems, answering machines, telephone sets, and PBXs. Before 1
Good point about NT. The original author of Windows NT, and I think still a leader in MS OS development, was David N. Cutler. He joined MS from DEC. He left DEC when they cancelled Mica, a VMS-like operating system that was decoupled from VAX hardware. Mica was being written at DECwest near Seattle, a site opened at Cutler's request -- he was that influential.
Before moving west, Cutler had been the lead author of VAX/VMS (1977), reportely writing a rather large share of the V1.0 kernel himself. (He was said to be very, very fast at assembly programming, but created inefficient code. Very "V1".) Before writing VMS, he had worked on RSX-11. The original VAX/VMS was upward compatible from RSX, and indeed included RSX as a mode. (By the mid-1980s, PDP-11 emulation capabilities had been dropped from VAX hardware.)
People who are familiar with the innards have said that NT is closely modeled on VMS (which was decoupled from the VAX when Alpha came out; it has since been ported to Itanic). But that's not a reason to whine. Cutler had some good ideas, and he expressed them several different ways. He almost certainly didn't lift DEC code directly into NT -- although at the time, DEC did indeed make most VMS sources readily available (on microfiche -- not to be too easy to use). Oh yeah, VMS was mostly written in assembler and BLISS, not C.
If Linus' efforts were unoriginal, how do Cutler's at Microsoft compare? Is the pot calling the kettle black? Personally I think neither is guilty of wrongdoing. I'm even getting to sort of like Win2k, which I use on the laptop, though XP's bloat is just annoying, and 98's instability is really annoying.
No, no shame. SpecOps was trolling for dollars by keeping quiet about the WINE code they're using. Now they're spinning.
So in this new letter, they're admitting that there's WINE in there, though not saying how much. And they're adding improvements atop it, which Codeweavers also does. And when it's released, per the GPL, they'll apparently release their modified source code where required. Okay. Whether David is useful or not remains to be seen.
> forward it through your business's mail server.
You're not reading the OP correctly. My business, like many small businesses, doesn't own a mail server. It owns a domain name, which works via a mail-forwarding service on mydomain. Outgoing mail goes via whatever ISP I happen to be on at the time; i.e., the one at home when I'm home, the one at the office when I'm at the office, or the one at the hotel when I'm at a hotel.
My clients don't care where my mail originated. They know it's from me. I'm not Paypal or BankOne.
If I had a mail server of my own, then another popular (but dumb) antispam measure, port 25 blocking, would cause it to be inaccessible from at least some locations.
Now it sounds like a bad idea for both semantic (what it does) and syntactic (how it is coded) reasons!
The syntactic bit is easy -- XML is hardly appropriate for a DNS function. Mickeysoft is running around patenting XML schemas, and it adds a new layer of complexity to DNS. But then bad syntax is usually dealt with by code.
The semantic bit is worse -- SPF doesn't block spam unless the mail system makes it mandatory, after all, so until 100% compliance is reached, non-SPF mail will still have to be accepted. But wait -- SPF doesn't block spam! It just blocks spam where the From: is not right. Spammers can still create new domains on a hit-and-run basis, and they'll pass SPF. So it's another blast-proof vault door stuck onto a grass hut, a silly waste of time. The only potential real benefit, I suspect, would be to make phishing harder. The address will have to be slightly different from the spoofed domain. But that leaves plenty of opportunity to create deceptively-close hit-and-run domains (like, say, pay-pa1-approva1.com).
Worse, of course, is the collateral damage. How will I be able to send mail using my own business' domain, as I do today, when it is going out via an ISP server? My "from" address is an alias, not a real sender, and I use it to send via more than one ISP, depending on where I am. SPF seems to make this a lot harder, thereby forcing more people to put their ISPs' name in the From: field, rather than their own. Since email is not portable, a user's address is lost when they change ISPs, or when their ISP changes names (mediaone->attbi->comcast). Personal domains (forwarded via a service like mydomain) solve this. Will SPF kill mydomain?
I repeat what I've said before. The only way to kill spam is to stop having all email be totally, absolutely, "free" of charge in any quantity. This is not the topic to discuss solutions, but they are certainly possible, and they aren't SPF.
Allowing "new competition to enter the arena" is often ridiculous, when there is a strong natural monopoly component to the service in question. Natural monopoly is an economic term for a product with a very high entry cost to produce and a large economy of scale. Telephone wire is a good example: If somebody has a 100% market share, and loses 20% to a competitor, and the cost is to run the wire past the customer site (which is the case), then the incumbent's unit cost will be 1/4 of the competitor's. This makes competition extremely difficult.
The Telecom Act recognized this, by requiring the incumbents to unbundle their facilities. The incumbents went along with the deal (getting other things in trade), but immediately tried to back out. It's bloody out there now on the telecom streets.
SBC, VZ, and the other incumbents have a natural monopoly on the local wire, even though it's no longer a de jure monopoly. There's no natural monopoly on Internet service, servers, information, etc., but if the monopolists can exert full control on content, as Gilder wants and MCI doesn't want them to, then all of those competitors will be shut out.
The original post is wrong. It implies that Gilder supports the MCI position. But Gilder's testimony opposes it. Gilder is, like many extreme right-wingers, a fan of unregulated monopolies. He does not believe that monopolies should be regulated; he thinks that technology will magically render them powerless. His record in picking technology investments, however, is rather spotty, to be charitable.
In Gilder's world, the incumbent telephone companies, who had government-granted monopolies, should be allowed to have total control over their usage and content. No competing ISPs, no uncensored web sites, no competing web merchants, if that's what the monopoly wants. If they want to charge $100/month plus $5/hour for dial-up access, fine. Anything else is, to him, excessive government regulation. He'd permit somebody else, of course, to string new wires on the street, but the impracticality of that is not his problem.
He should be relegated to the dumpster of other failed nut case prophets, and left with his fellow creationists to ponder the problems of a world where scientists and rational thinkers are allowed to question his faith.
Just to be more specific about the history...
Intel licensed AMD to produce their designs, as a second source, up through the 80286. Intel masks and all. By the time the 386 came out, Intel didn't need AMD any more (they had multiple fabs and a good enough reputation, plus a lock on PC-compatible chips). So they told AMD that the agreement didn't apply any more. I don't remember if AMD won or lost on the 80386. But it certainly didn't last until the 486. So AMD did their own design, without any help from Intel. The court did note that a number could not be trademarked. It was thus never the "80486"; I think "i486" was a trademark, not that anybody cared, and that's why the next Intel chip was "Pentium".
AMD's "586"-class chip, the K5, was a dog. They then bought NexGen and adapted its RISC-innard design to the K6, which rocked, and fit a Pentium socket. Intel put tighter patents on the PII socket so AMD built the Athlon on DEC's Alpha socket electrical design.
Intel didn't have to change the ISA (drop the NX, for instance) in order to be legal. Either they goofed, or they sabotaged their own 64-bit x86 upgrade (as others here have suggested) in order to create a niche for the Itanic.
Lexmark obviously makes the Xerox printer in question. They design in little differences, like that ink clip, in order to create separate ink streams, which are where the money is. The article shows how one person got around one such difference.
Dell printers are also Lexmark; I don't know if they differentiate on ink. But if you read the printer reviews in the PC mags, you'll see the estimated per-page ink prices. Guess whose are highest? Lexmark ink is much costlier than the other major brands (HP, Canon, Epson), with Dell-branded ink usually a notch higher yet.
Good point -- Mandrake ADDS VALUE via its own proprietary printer-installation routine. Maybe it's 99% CUPS, maybe it's 99% their own, but at least I got it to work. Fedora Core sounds like it's running raw CUPS. I couldn't imagine Red Hat going out of their way to make it that bad!
And Kudos to ESR on this piece in general. CUPS is just one example of the abuse of GUIs across Linux. The X Windows configuration tools in some distros are even worse. When you use a GUI that sets some value or other, the first rule is that the value displayed should be the CURRENT value. After all, it's a tool for seeing, as well as setting. But no, in some Linux tools, the GUI always comes up with some weird default value! It ignores what's there and wants ohhh so badly to overwrite it with something else. What an awful thing to do!
That's PRECISELY the kind of UI polishing that Linux needs. Not a few more sci-fi desktop themes.
AMD's big problem isn't marketing, it's manufacturing. AMD is doing a pretty good job with the resources they have, but Intel has much larger capacity. Nobody fabs like Intel, and nobody has fabs like Intel. So if AMD had more orders, they'd just have bigger backlogs. They may be able to get some third-party fabrication, like from IBM, but Intel would not let its fabs sit idle. I prefer AMD chips myself, at least for desktop-type applications (Intel still wins on mobile chips), but this is not like OS/2 (which I also preferred in its heyday) -- chipmaking is expensive!
The article, which was truly stupid, ran on ZDNet. While that sounds like Ziff-Davis, it's not. ZDNet is part of CNET, a separate company; they bough it from Ziff-Davis several years ago. Ziff-Davis still puts out PC Magazine (pcmag.com) and Extreme Tech, among other things. Sort of like how Lycos owns Wired.com, not Wired magazine, though those two still have a working relationship that is not apparent between ZDNet and CNET.
Among CNET properties, ZDNet seems to be oriented towards Windows-loving IT pros. Maybe that's part of its Ziff-Davis heritage, or just a marketing decision.
I've posted this before, when the topic came up on another occasion, but it's worth repeating.
The FCC is NOT going to regulate computer-to-computer "phone" calls. If you run voice over your Internet connection, as an application, it's your business, and that's that. Even the guy who drafted the infamous ACTA petition in 1996 now thinks VoIP is cool stuff.
The problem is the phone call between the consumer with a plain old phone line and the VoIP network. "Phone to phone" and "phone to computer" calls have a telco leg that's just a plain old voice call. Under current law, a phone call can be either "telephone exchange service" or "exchange access service". The former is basically taken to mean a local call, though the legal definition is a bit more expansive. The latter is taken to be the local phone company's leg of a toll call (what AT&T or MCI buys). Guess which one costs more.
Now if all VoIP calls were treated as local ("telephone exchange service"), then the local telephone companies (think: Bells) would lose money that they now make from exchange access service ("switched access"). And the rural phone companies, who charge the long distance companies MUCH more than the Bells for that service, in order to compensate for higher costs (that is, to subsidize local service to the sticks), are very protective of switched access revenues. And the flyover states each have two senators.
So the main issue will come up around the far end of a Vonage call, for instance -- if Vonage is a long-distance company, they will have to pay access when they deliver a long distance call. Just like other long distance companies. Skype's on-net calls, and FWD, won't be touched as long as they are on net. Count on it.
Ideally, the whole access thing would go away, and the distinction between access and local would be moot. That's the way it works in msot of Europe, I think -- it's an American tradition to classify things to death, and let the lawyers litigate like crazy over the classification. How many billable lawyer hours do you think this case will be worth in Washington?
Those are interesting, thank you for posting them.
Both, however, seem dependent upon having both sides of the conversation join. In the case of Mailbox Reputation Network, it seems to be a voucher-based (as in I vouch for you, a web of trust) system. In the case of sudonames, it seems to be a whitelist system, where non-members can't send mail to you.
If both were universal, they'd probably work, especially sudonames, because that seems to count the credits, making host-hijacking worthless (it would run out of credits, which wouldn't happen right away with MRN, as I see it). But if I joined sudonames today, random people trying to reach my sudonames address would get bounced.
But then any micropostage scheme could have that problem, even mine, though a whitelist would exempt recipients from needing credits (stampettes).
I'm drowning in spam, and it's getting in the way of my job. The only solution that can possibly work is one that involves putting a price tag on spam. So here's my proposal (which I've put on here before, btw; this is not a new topic). The only way to put a price tag on spam is to put a price tag on email. But it doesn't have to apply to all email.
The price, then, is for the right to touch MY mailbox IF you're a stranger -- if you're a mailing list that I've subscribed to, you would go onto my whitelist, and come in postage-free. If you are somebody I know, you go onto my whitelist, and come in postage-free. Yes, for this to work, there has to be some way for the POP server (NOT the client) to maintain per-user whitelists.
If you're not on my whitelist, you need to use a one-time "stampette", whose price would have to be high enough to discourage spammers, but low enough to not bother anybody worthwhile. I'm thinking around a quarter-cent per message, but it wouldn't be fixed by anyone in particular. These stampettes would be issued on a free-market basis, and anyone could set up a micropostage service, provided that the *recipient* whitelisted it. So if somebody were giving away stamps at, oh, a million per dollar, then spammers would use them, and those stamps wouldn't be on my whitelist. Again, it's a free market solution, no government intervention.
ISPs, in this scheme, should issue all subscribers a batch of stampettes (which mail clients would learn quickly to attach, if needed). A thousand for a quarter-dollar (or quarter-Euro) would be more than enough for a month, don't you think? How many strangers (or first-time correspondents) do you write to?
The point is: IPv4 is not the problem; IPv6 is not a good solution to the alleged problems.
Not all of the IPv4 address space has been parceled out. ICANN has a lot of the original "Class A" space available. The space from 65-126 was never allocated as Class A; it has been parceled out to the three worldwide number assignment bodies on a demand basis. Quite a bit is left. There are also some Class As that can probably still be reclaimed, in whole or in part. NAT has also helped a lot in holding down demand for numbers.
Even if numbers were in desperately short supply and IPv4 couldn't handle the job, IPv6 wouldn't be the answer. It's plug-ugly, the bastard child of two amateurish hacks by IETF insiders (Steve Deering's SIP -- the current SIP is at least the second holder of the name -- and Paul Francis/Tsuchiya's PIP), melded together sloppily in order to get "consensus". IAB had already accepted TUBA, a far cleaner solution, but Vint changed his vote. What a friggin' disaster. TUBA (TCP and UDP over a CLNP profile) had already been implemented on all of the major routers of the day. It just hit a wall of "NIH", since its creation was tained by its OSI connection.
But that was all before the Internet was big or open. If a replacement for v4 were really needed, it should not be yet another old hack. It should be something built with today's requirements in mind, not 1990's. Real network research, alas, seems to have shut down at about the time that the Internet became commercially important. Too valuable to question, I suppose, and all the newcomers from Microsoft to SCO must imagine that it must have been well thought out in the first place (hah! it was government research, still alpha or maybe early beta work in progress) but that is a terrible way to maintain it.
And back to our Asian friends: Software has never been their strong point. Nor has questioning authority; too many, I suspect, assume that the TCP/IP suite has too much authority behind it. Asia's marvelous at mass-producing hardware, an art which involves being able to reproduce things perfecty in media that make it difficult to do so. So if they accept IPv6, it is not necessarily proof that it's, say, Toyota-grade technology. Even Japan has its clunkers. Remember Pink Lady and Jeff?)